Republic of Moldova: Selected Issues

International Monetary Fund

doi:10.5089/9781451825046.002.A003

Jump to Content Jump to Main Navigation

Republic of Moldova: Selected Issues

Author: International Monetary Fund

Publication Date:: 17 Feb 2005

eISBN:: 9781451825046

Language:: English

Keywords:: ISCR; CR; enterprise; government; migrant; equilibrium exchange rate; REER appreciation; capital stock; nominal exchange rate; world price; Real effective exchange rates; Real exchange rates; Exchange rates; Wages; Purchasing power parity

Download PDF (329.6 KB)

This Selected Issues paper analyzes the macroeconomic impact of workers’ remittances on Moldova. The paper focuses on Moldova’s labor emigration since the late-1990s using survey data designed to shed light on the economic and social consequences of migration. The survey results are broadly consistent both with the findings from balance-of-payments data and with the stylized facts in the labor migration literature. The paper also examines various indicators to assess the appropriateness of the current exchange rate level.

Abstract

III. External Competitiveness³⁰

A. Introduction

92. Badly hit by the 1998 regional crisis, Moldova’s economy did not return to positive rates of growth until 2000. The pace of recovery picked up in subsequent years, with GDP growth averaging between 5 and 7 percent annually. However, the underlying domestic consumption boom and related increases in merchandise imports have given rise, over time, to concerns about the sustainability of the recovery.

93. Importantly, recent trends have raised questions about Moldova’s external competitiveness in general, and the appropriate level of the leu exchange rate in particular. With wages a fraction of the European Union average, Moldova has traditionally been regarded as a low-cost country, with a large potential—given right policies and development of a business-friendly environment—to attract direct foreign investment and boost exports. However, the steady appreciation of the Moldovan leu against the dollar, fast wage growth, and a dramatic widening of the trade deficit (from 23 to 32 percent of GDP between 2002 and 2003), have attracted policymakers’ attention (Figure 1).

94. This chapter assesses Moldova’s external competitiveness in light of these recent developments. We use several approaches to estimate the appropriate level of the exchange rate: (i) purchasing power parity (PPP); (ii) constant equilibrium real effective exchange rate (REER); (iii) equilibrium wage from a panel regression; (iv) equilibrium wage from a production function; and (v) equilibrium REER in a small open economy, three-good model. We conclude that Moldova does not face immediate risks of losing external competitiveness. Based on a number of indicators, its exchange rate appears undervalued, and is likely to continue to rise in real effective terms.

95. While the REER appreciation in recent years has been in line with the experience of other transition countries, it has been reinforced by several factors specific to Moldova. Particularly relevant are the exodus of working-age population, and the related large and growing inflows of migrants’ remittances. Emigration has helped raise the actual wage (by reducing unemployment) and the equilibrium wage (by increasing the capital/labor ratio in the economy). At the same time, remittances have boosted national disposable income and domestic demand, leading to price increases in nontraded goods.

96. Going forward, maintaining and improving external competitiveness in Moldova is inextricably linked to the overriding policy challenge of accelerating economic development. In that sense, competitiveness needs to be viewed not only in terms of safeguarding external sustainability, but also as Moldova’s ability to build solid economic growth by attracting much-needed foreign direct investment.

B. Estimating the Equilibrium Exchange Rate

(i) Purchasing Power Parity

97. The conventional approach to estimating a currency’s equilibrium level is based on some form of purchasing power parity (PPP). This consists in constructing an index R, defined as the nominal exchange rate (E), multiplied by the ratio of an aggregate index of world prices (P^*) divided by an index of domestic prices (P):

R = E \cdot \frac{p *}{P}

98. In its most restrictive form, the equilibrium level of E is assumed to be that which equalizes the foreign and domestic price levels, expressed in a common currency. In other words, one that makes R equal to unity. A well-known example of this index is the Economist’s McDonald’s Big Mac standard: an over/undervaluation of a currency is calculated by comparing (i) the currency’s dollar exchange rate that would obtain from equalizing the domestic price of McDonald’s Big Mac to its price in the United States, and (ii) the actual dollar exchange rate. According to this index, the Moldovan leu was undervalued by 33 percent in May 2004.³¹ The same idea lies behind the PPP-constructed exchange rates calculated for a broader set of goods and services. For example, based on the WEO data of PPP, the Moldovan leu was undervalued by 74 percent in 2003.

(ii) Constant Equilibrium Real Effective Exchange Rate

99. Under a less constraining approach, the equilibrium value of R is allowed to diverge from unity, but remains constant at some level, at which the balance of payments is deemed to be in equilibrium.³² For Moldova, it is difficult to select an appropriate reference year, given its tumultuous recent economic history, which can be traced in the movements of the leu real exchange rate against its main trade partners (Figure 2). After 1992, the country went through a difficult period of post-communist transition, characterized by structural changes that affected both its domestic economy and the territorial composition of its external trade (Figure 3). Later, it was hit hard by the 1998 regional crisis, which led to a collapse in exports in 1998 and 1999, and a forced reduction in imports. A measure of stability returned with the resumption of economic growth in 2000.

100. Some considerations favor 2001 as a possible reference year. The current account deficit narrowed to 2.5 percent of GDP; and both gross foreign exchange reserves and the REER remained relatively stable. In 2002, the REER—as well as the leu/euro and the leu/ruble exchange rates—began a decline that was not reversed until April 2003. Since then the REER appreciated markedly, and by November 2004 it was close to its 2001 average value. On this basis, it could be concluded that, in October 2004, the leu was at about its equilibrium level.

(iii) Equilibrium Wage Based on a Panel Regression

101. The idea that the real exchange rate will tend to return to a previously established, unchanging equilibrium level, is not universally accepted in the economic literature.³³ The concerns are particularly germane when it comes to post-communist countries undergoing a transition to market economy with all attendant structural changes and relative price realignments. For example, substantial real exchange rate appreciations have been documented for CEE countries,³⁴ explained by: (i) the Balassa-Samuelson effect; (ii) growing incomes in the domestic economy, which lead to higher spending and a rise in the price of nontradables; (iii) recovery from the initial undervaluation. Considerations (i) and (ii) point to a relation between a country’s income level and the extent of its currency “undervaluation” relative to its currency PPP level. This can be observed in Figure 4, which suggests that the leu was undervalued by about 74 percent, as mentioned in paragraph 7. One might therefore attempt to estimate the equilibrium exchange rate as a function of selected underlying fundamentals—including the level of average per capita income. As these fundamentals change over time, so will the equilibrium exchange rate.

102. An innovative approach used by Tiffin (2004) is to pool transition economies’ data through a cross-section panel to estimate the equilibrium real exchange rate. The level of dollar wages in the manufacturing sector is used as a measure of the real exchange rate for the countries in question. The attractiveness of this approach lies in the fact that (a) manufacturing goods are well represented in international trade; (b) labor cost plays an important role in determining competitiveness; and (c) data on manufacturing wages are available and comparable (compiled by the International Labor Organization in a standardized format).

103. The model is set up to estimate the equilibrium dollar wage in manufacturing as a function of selected “fundamentals”: PPP-adjusted GDP per capita; share of agriculture in GDP; and the gross secondary-school enrollment rate. These variables serve as productivity proxies. The equilibrium wage estimated on the basis of these variables represents the wage that a country can afford based on its capital stock and overall level of development. Tiffin finds that, for CEE countries that recently acceded to E.U. membership, as well as for those of the Former Soviet Union (FSU), prices and wages are below the level that would reflect their true economic value. However, this undervaluation has been declining over time. In CEE countries, dollar wages approached their equilibrium values by the late 1990s. In FSU countries, the undervaluation has been more resistant, partly reflecting the overshooting following the 1998 regional financial crisis.

104. Moldova’s experience reflects broadly that of other countries (Figure 5). With the actual wage rising for most of the 1990s until 1998, the undervaluation gap narrowed, before widening sharply as a result of the 1998 crisis. Between 2000 and 2004, the gap was cut in half—from 52 to 25 percent—as the equilibrium wage increased by 26 percent and the actual wage doubled. The most recent data suggest that Moldova’s exchange rate has still room to appreciate. Looking ahead, the experience of “advanced reformers” may serve as a guide to the likely future path of the real exchange rate of the Moldovan leu. Assuming economic growth takes firm hold in Moldova, the equilibrium wage would be expected to continue to rise. That, in combination with the gradual movement of the actual wage toward equilibrium, implies a faster rate for actual wage growth as well as for the real exchange rate appreciation.

(iv) Equilibrium Wage Based on a Production Function

105. In the preceding formulation, the equilibrium wage determinants were variables chosen as proxies for the capital stock and overall level of development. Alternatively, one could attempt to model changes in the equilibrium wage as a function of changes in the capital stock or in the capital-labor ratio—based on a conventional cost-minimization framework with a two-factor production function (Box 1).

Box 1.

Estimating the Equilibrium Wage Based on the Capital/Labor Ratio

Assume a Cobb-Douglas production function

\begin{matrix} Y = A \cdot L^{α} \cdot K^{1 - α} & (1) \end{matrix}

where: Y = output

L = labor
K = capital
A, α = parameters

We obtain marginal products of labor and capital by differentiating (1) with respect to L and K:

\begin{matrix} \frac{d Y}{d L} = α \cdot A \cdot {(\frac{K}{L})}^{1 - α} = α \cdot A \cdot k^{1 - α} & (2 a) \end{matrix}

\begin{matrix} \frac{d Y}{d K} = (1 - α) \cdot A \cdot {(\frac{K}{L})}^{- α} = (1 - α) \cdot A \cdot k^{- α} & (2 b) \end{matrix}

where k = K/L

In equilibrium, the wage (w) will equal the marginal product of labor, hence from (2a) we can write:

\begin{matrix} w = α \cdot A \cdot k^{1 - α} & (3) \end{matrix}

By differentiating (3), we obtain

$d w = α \cdot A \cdot (1 - α) \cdot k^{- α} d k,$ and dividing through by w we get a percent change in w (=dw/w):

\begin{matrix} \frac{d w}{w} = (1 - α) \cdot \frac{d k}{k} & (3 b) \end{matrix}

Differentiating k, we can write also

\begin{matrix} \frac{d k}{k} = \frac{d K}{K} - \frac{d L}{L} = \frac{I}{Y} \cdot \frac{Y}{K} - δ - \frac{d L}{L} & (4) \end{matrix}

where: I = gross investment; 5 = rate of depreciation of physical capital

By substituting (4) into (3b), we obtain

\begin{matrix} \frac{d w}{w} = (1 - α) \cdot (\frac{I}{Y} \cdot \frac{Y}{K} - δ - \frac{d L}{L}) & (5) \end{matrix}

This implies that the changes in equilibrium wage will depend positively on the investment/output ratio, and inversely on the capital/output ratio, depreciation rate and labor force growth.

106. Using data for investment and labor force for the period 1995-04, we estimate that Moldova’s capital/labor ratio increased by 28 percent. Roughly two-thirds of this increase were due to the decline in the denominator (labor force), and one-third due to an increase in the capital stock through new investment (Table 1).

Table 1.

Moldova: Capital/Labor Ratio, 1995-04

(Cumulative change, percent, unless noted otherwise)

Sources: Moldovan authorities; and staff calculations.

107. Labor force data and capital stock estimates need to be treated with caution. The true extent of labor emigration is not known, and capital stock data are not directly available, but have to be estimated from annual investment flows using heroic assumptions. The paucity of data allows only to estimate the change in the equilibrium wage, rather than its level (cf. equation 5 in Box 1). In Figure 6, the average level for 1995-2004 is calibrated to correspond to the average level estimated by Tiffin for this period. The exercise provides supporting evidence for the rising equilibrium wage in Moldova: the increase in the capital/labor ratio, and the implied increase in the equilibrium wage, are in line with Tiffin’s estimates obtained through a panel regression.

(v) Equilibrium Real Exchange Rate in a Small Open Economy, Three-Good Model

108. The single most important defining characteristic of Moldova’s economic development in recent years has been the fast growth in remittances from abroad.³⁵ Unfortunately, none of the approaches outlined in the preceding discussion is equipped to model the equilibrium REER as a function of changes in sustainable income and capital flows in the balance of payments. It may be therefore useful to consider a model that would be better suited to analyze structural adjustments of the kind undergone by a small country receiving large amounts of foreign exchange inflows.

Box 2.

Outline of a Three-Commodity Model^1/

A country produces two goods—a nontraded domestic good, D, and an export good, X. It consumes two goods—the domestic good and the imported good, M. The corresponding prices are Pd, Px, and Pm. Goods D and X are assumed to be imperfect substitutes in production—a characteristic captured by the economy’s production possibility frontier, specified as a constant elasticity of transformation (CET) function. Profit maximization by producers implies that the relative supplies of D and X depend on their relative prices, Pd and Px, and on the elasticity of transformation, Ω. Goods D and M are assumed to be imperfect substitutes in consumption, with a constant elasticity of substitution (CES) function. The first-order condition for utility-maximizing consumers implies that relative demands for M and D will depend on their relative prices, Pm and Pd, and on the elasticity of substitution, σ. The domestic prices of the two traded goods (M and X) equal their world prices (π^m and π^x, respectively) times the nominal exchange rate (E). The world prices are exogenous (small country assumption). Finally, the balance of trade constraint states that the sustainable trade balance (exports minus imports) is set exogenously.

The model can be reduced to three equations:

\begin{matrix} \frac{M}{D} = c_{1} \cdot {(\frac{p d}{p m})}^{σ} & (1) \end{matrix}

\begin{matrix} \frac{X}{D} = c_{2} \cdot {(\frac{p x}{p d})}^{Ω} & (2) \end{matrix}

\begin{matrix} π^{m} \cdot M = λ \cdot π^{x} \cdot X & (3) \end{matrix}

where the constant terms in equations (1) and (2) represent the share parameters from the CES and CET functions. Parameter λ in equation (3) is the country’s sustainable balance of trade, or the proportion by which imports can exceed exports.

By log differentiation, where $d \log (X) = \hat{X} = \frac{d X}{X},$ we obtain

\hat{M} - \hat{D} = σ \cdot (\hat{P} d - \hat{P} m) ......................... (1 b)

\hat{X} - \hat{D} = Ω \cdot (\hat{P} x - \hat{P} d) .......................... (2 b)

{\hat{π}}^{m} + \hat{M} = \hat{λ} + {\hat{π}}^{x} + \hat{X} ......................... (3 b)

The nominal exchange rate, E, is chosen as the numeraire, so that Ê = 0 and $\hat{P} m = {\hat{π}}^{m}$ and $\hat{P} x = {\hat{π}}^{x}$ . As the world prices are set exogenously, the only endogenous price in the model is the price of the domestic good (Pd), which also determines the real exchange rate (R). Solving for the real exchange rate, we obtain

\hat{R} = \hat{E} - [\hat{P} d - \frac{(σ \cdot {\hat{π}}^{m} + Ω \cdot {\hat{π}}^{x})}{σ + Ω}] = (\frac{{\hat{π}}^{m} - {\hat{π}}^{x}}{σ + Ω}) - \frac{\hat{λ}}{σ + Ω} ...................... (4)

Equation (4) shows the real exchange rate—defined as the nominal exchange rate, adjusted for the inflation differential between the home country and its trading partners—as a function of the two right-hand-side terms: the terms of trade; and the sustainable balance of trade. Equation (4) makes it clear that the conventional approach—based on the assumption that there is some unchanging equilibrium level for the real exchange rate—is valid only if the two terms on the right-hand side are equal to zero: i.e., there is no change in the country’s terms of trade, or in the sustainable level of foreign income or capital inflows.

^1/ Based on Devarajan et al. (1993).

109. Box 2 outlines a model of a small, open economy with two tradable goods (export good and import good) and a nontraded good. Equation (4) suggests that the equilibrium REER changes in response to shifts in the terms of trade and in the sustainable income and capital flows. Between 1996 and 2004, Moldova benefited from a modest improvement in its terms of trade and a large increase in workers’ remittances, which have been financing a progressively larger portion of its trade deficit (Figure 7).

110. The model suggests that Moldova’s equilibrium REER has increased in response to the rising inflows of remittances. Figure 8 shows the cumulative appreciation since end-1996, with the contribution from the change in sustainable trade balance corresponding to the influence of remittances inflows.³⁶ Under the assumption of elasticities of substitution and transformation equal to 0.75, the model estimates a cumulative appreciation of about one third in Moldova’s REER between 1996 and mid-2004.

111. Admittedly, the estimated impact will depend on the values selected for the parameters of the model. To gauge the robustness of the model predictions, the results were calculated for alternative values of elasticities of transformation and substitutions (Table 2). They suggest a range for the increase in the equilibrium REER of 23 to 51 percent between 1996 and 2004.

Table 2.

Moldova: Cumulative Change in Equilibrium REER

(Since 1996, in percent)

Sources: Moldovan authorities; and staff calculations.

Table 2.

Moldova: Cumulative Change in Equilibrium REER

(Since 1996, in percent)

Transformation elasticity, Ω	Substitution elasticity, σ
Transformation elasticity, Ω	0.50	0.75	1.00
0.50	51	39	32
0.75	39	32	27
1.00	32	27	23

Sources: Moldovan authorities; and staff calculations.

Table 2.

Moldova: Cumulative Change in Equilibrium REER

(Since 1996, in percent)

Transformation elasticity, Ω	Substitution elasticity, σ
Transformation elasticity, Ω	0.50	0.75	1.00
0.50	51	39	32
0.75	39	32	27
1.00	32	27	23

Sources: Moldovan authorities; and staff calculations.

112. Figure 9 contrasts the predicted range for the cumulative appreciation of the equilibrium REER with the actual cumulative change in REER since end-1996. It suggests that the cumulative increase in the actual REER (10 percent) was below the increase estimated by the model for the equilibrium REER.

C. Conclusion

113. The preceding analysis reveals that the Moldovan leu is unlikely to be overvalued at present. According to the various approaches used here to calculate the equilibrium REER, the leu remains below its equilibrium value despite the REER appreciation observed in 2003 and 2004. The extent of undervaluation varies according to the method used, but its mid-point could be put in the 20 to 40 percent range. While the increase in dollar wages and the nominal appreciation of the leu—particularly against the dollar—have captured public attention, some aspects have received less emphasis: (i) the large undervaluation of the REER following the 1998 regional crisis, implying a large initial gap between the actual and equilibrium REER; and (ii) the role of balance of payments inflows (remittances) in raising the equilibrium REER.

114. Further REER appreciation is likely. Absent other developments, the REER has some room to appreciate, and will likely do so if Moldova’s economy operates close to its full potential, the labor market remains tight, and remittances continue to grow. Rather than viewing the leu appreciation as harmful to the economy, our analysis suggests that, by raising the returns in the nontraded sector relative to those in the traded sector, it could facilitate a reallocation of resources in the economy that could lay the foundation for long-term growth. With remittances providing an important portion of balance of payments financing, domestic production could be redeployed toward more goods and services needed for domestic investment. The hitherto neglected domestic infrastructure—transportation network, electricity transportation and distribution, communications—could benefit from this reallocation of resources. Over time, the modernization of the domestic infrastructure would help strengthen the economic potential by helping raise the economy-wide productivity growth.

115. Understanding the reasons behind the real leu appreciation and its implications is important for selecting appropriate policy responses. Resisting nominal leu appreciation will be ineffective as a way of containing real exchange rate appreciation—which will be instead obtained through higher inflation—and will be undesirable if the real exchange rate appreciation reflects changes in its equilibrium value. Our analysis also serves as a reminder that, in the long run, Moldova’s competitiveness needs to be viewed as the ability of the economy to generate increases in incomes through higher investment and productivity growth. This, in turn, underscores the need to focus attention on structural impediments to higher private investment.

References

Devarajan, S., J. Lewis and S. Robinson, 1993, “External Shocks, Purchasing Power Parity and the Equilibrium Real Exchange Rate”, World Bank Economic Review, 7 (1). (Washington: World Bank).
- Search Google Scholar
- Export Citation
Halpern, Laszlo, and Charles Wyplosz, 1997, “Equilibrium Exchange Rates in Transition Economies,” IMF Staff Papers, Vol. 44, pp. 430–61 (Washington: International Monetary Fund).
- Search Google Scholar
- Export Citation
Krajnyák, Kornélia, and Jeromin Zettlemeyer, 1998, “Competitiveness in Transition Economies: What Scope for Real Appreciation,” IMF Staff Papers Vol. 45, pp. 309–362 (Washington: International Monetary Fund).
- Search Google Scholar
- Export Citation
Rogoff, Kenneth, 1996, “The Purchasing Power Parity Puzzle,” Journal of Economic Literature, Vol. 34, pp. 647–68.
- Search Google Scholar
- Export Citation
Schadler, Susan, Paulo Drummond, Louis Kuijs, Zuzana Murgasova, and Rachel van Elkan (2004), Adopting the Euro in Central Europe. Challenges of the Next Step in European Integration, IMF Occasional Paper 234 (Washington: International Monetary Fund).
- Search Google Scholar
- Export Citation
Tiffin, Andrew (2004), “Competitiveness, Convergence, and the Equilibrium Real Exchange Rate,” in Ukraine: Selected Issues, Country Report No. 05/20 (Washington: International Monetary Fund).
- Search Google Scholar
- Export Citation

Prepared by Milan Cuc.

The latest available data.

This sustainable level could be different from zero. For example, with a growing GDP, a country could run a current account deficit that would leave the stock of foreign liabilities constant as a percentage of GDP.

The tendency of a variable to return to some average, constant value has been termed “mean reversion”. Rogoff (1996) documents a slow reversion for exchange rates.

See, for example, Schadler et al. (2004).

See Chapter I.

See equation (4) in Box 2.

Other Publishers

Save
Cite
Email this content

Share Link

Copy this link, or click below to email it to a friend
Email this content
or copy the link directly:

https://www.elibrary.imf.org/view/journals/002/2005/054/article-A003-en.xml
The link was not copied. Your current browser may not support copying via this button.

Link copied successfully